THE TENDENCY OF ORGANIZATIONAL RESEARCHERS, EVALUATION SPECIALISTS, POLICYMAKERS, AND PROGRAM DEVELOPERS TO IGNORE THE ISSUE OF PROJECT IMPLEMENTATION IS NOTED. THE CONCEPTUAL FRAMEWORK OF THE MICHIGAN STUDY, WHICH EMPLOYED RESEARCH QUESTIONS ABOUT ENVIRONMENTAL INFLUENCES ON PROJECT IMPLEMENTATION, EFFECTIVENESS, AND INSTITUTIONALIZATION, IS OUTLINED. PRELIMINARY FINDINGS ARE REPORTED FROM INTERVIEWS WITH POLICE PROJECT PERSONNEL REGARDING THE DEVELOPMENT OF EACH PROJECT, ITS ACTUAL OPERATION, THE SOCIAL AND POLITICAL CONTEXT OF THE PROJECT, AND THE FACTORS THAT APPEARED TO FACILITATE OR HINDER THE SUCCESS OF THE PROJECT OR ITS INSTITUTIONALIZATION. FOUR PRIMARY ISSUES IN THE IMPLEMENTATION PROCESS -- PROJECT GOAL AND OBJECTIVE CLARITY, GOAL CONSENSUS, INTERDEPENDENCE OF VESTED INTERESTS, AND LOCAL MOTIVATIONS FOR OBTAINING AND USING FEDERAL SUPPORT -- ARE CONSIDERED. THE EVALUATION FINDINGS IMPLY THE FOLLOWING: (1) PROBLEMS CREATED BY GOAL AMBIGUITY AND LACK OF GOAL CONSENSUS AMONG RELEVANT INDIVIDUALS AND ORGANIZATIONS SHOULD BE DEALT WITH THROUGH MORE INTENSIVE SITE PREPARATION THAN IS CUSTOMARY FOR MOST DEMONSTRATION PROJECTS; (2) EARLY PHASES OF IMPLEMENTATION SHOULD EMPHASIZE FORMATIVE, AS OPPOSED TO SUMMATIVE, EVALUATION SO THAT FEEDBACK RELEVANT TO THE DEVELOPMENT OF THE PROJECT WILL BE PROVIDED; (3) EVALUATION OF THE IMPLEMENTATION PROCESS IS NECESSARY IF PROJECT RESULTS ARE TO BE GENERALIZED; AND (4) EMPHASIS ON THE IMPLEMENTATION PROCESS ULTIMATELY WILL ENHANCE THE ABILITY TO DRAW CAUSAL INFERENCES FROM PROJECT EVALUATIONS. A LIST OF REFERENCES AND READINGS IS INCLUDED.